Rethinking Collapsed Variational Bayes Inference for LDA
Rethinking Collapsed Variational Bayes Inference for LDA
复制标题
DOI：
--
复制
发表时间：
2012-06
期刊：
影响因子：
--
通讯作者：
Issei Sato;Hiroshi Nakagawa
中科院分区：
文献类型：
--
作者：
Issei Sato;Hiroshi Nakagawa
我们提出了一种新的解释与零阶泰勒展开近似的崩溃变分贝叶斯推理，称为CVB0推理，对于潜在的狄利克雷分配（LDA）。我们利用α-散度澄清了CVB0推理的性质。我们证明CVB0推理由两个不同的散度投影组成：α = 1和-1。这种解释将有助于阐明CVB0的作品。
We propose a novel interpretation of the collapsed variational Bayes inference with a zero-order Taylor expansion approximation, called CVB0 inference, for latent Dirichlet allocation (LDA). We clarify the properties of the CVB0 inference by using the α-divergence. We show that the CVB0 inference is composed of two different divergence projections: α = 1 and -1. This interpretation will help shed light on CVB0 works.